Method of Simulating Impact Events in a Multi-processor Computer System

• US 20100161296A1
• Filed: 12/24/2008
• Published: 06/24/2010
• Est. Priority Date: 12/24/2008
• Status: Active Grant

• Current Assignee: Ansys Incorporated
• Original Assignee: Livermore Software Technology Corporation
• Inventors: Wainscott, Brian
• Granted Patent: US 8,798,973 B2
• Time in Patent Office: Days
• US Class Current: 703/6
• CPC Class Codes: G06F 30/23 using finite element methods [FEM] or finite difference methods [FDM]

Abstract

A computer-implemented method of simulating an impact event in a finite element analysis used for assisting users to design or improve one or more structures is described. The structures are represented in a finite element analysis model that is divided or partitioned into a plurality of domains. Efficiency of the method is achieved when used in a computer system having multiple processing units and multiple contact interfaces defined and specified by users (engineers and/or scientists). Each domain is associated with or assigned to one of the processing units. A “group-able” correlation is established or determined between domains and contact interfaces so that data communications can be conducted in most efficient manner, for example, minimizing idle processing units during data communications.

Claims

1. A computer-implemented method of simulating an impact event in a finite element analysis used for assisting users to design or improve one or more structures, the method comprising:

receiving a finite element analysis model of one or more structures subject to an impact event in a computer system with a plurality of processing units coupled thereon, wherein the finite element analysis model contains a plurality of contact interfaces;

dividing a finite element analysis model into a plurality of domains containing at least a first domain and a second domain in the computer system, the first domain is associated with a first processing unit while the second domain is associated with a second processing unit of the processing units;

determining, in the computer system, a group-able correlation between each of the contact interfaces, and one or more of the domains;

performing, in the computer system, a time-marching simulation of the impact event having a plurality of solution cycles;

at each of the solution cycles,

creating and sending a first data communication message from the first processing unit to the second processing unit in accordance with the group-able correlation such that maximum possible data communications relevant to the contact interfaces are exchanged;

performing a plurality of computations for the respective contact interfaces of the finite element analysis at each of the first and second processing units with data received in the first data communication message;

creating and sending a second data communication message in a reverse direction of the first data communication message, the second message is configured to transmit data resulted from the computations; and

displaying results of the simulation in a monitor coupling to the computer system after all of the solution cycles have been computed.

2. The method of claim 1, wherein each of the one or more structures comprises an automobile.

3. The method of claim 1, wherein the impact event comprises a car collision.

4. The method of claim 1, wherein the computer system contains at least two processing units.

5. The method of claim 4, wherein the contact interfaces comprises at least two interfaces.

6. The method of claim 1, wherein the computer system comprises at least one computer, each having more than one central processor.

7. The method of claim 1, wherein the computer system contains more than one computer, each having at least one central processor.

8. The method of claim 7, wherein the central processor contains at least one processing core.

9. A computer-readable storage medium containing instructions for controlling a computer system to perform an impact event simulation in a finite element analysis used for assisting users to design or improve one or more structures by a method comprising:

receiving a finite element analysis model of one or more structures subject to an impact event in the computer system with a plurality of processing units coupled thereon, wherein the finite element analysis model contains a plurality of contact interfaces;

dividing a finite element analysis model into a plurality of domains containing at least a first domain and a second domain in the computer system, the first domain is associated with a first processing unit while the second domain is associated with a second processing unit of the processing units;

determining, in the computer system, a group-able correlation between each of the contact interfaces, and one or more of the domains;

performing, in the computer system, a time-marching simulation of the impact event having a plurality of solution cycles;

at each of the solution cycles,

creating and sending a first data communication message from the first processing unit to the second processing unit in accordance with the group-able correlation such that maximum possible data communications relevant to the contact interfaces are exchanged;

performing a plurality of computations for the respective contact interfaces of the finite element analysis at each of the first and second processing units with data received in the first data communication message;

creating and sending a second data communication message in a reverse direction of the first data communication message, the second message is configured to transmit data resulted from the computations; and

displaying results of the simulation in a monitor coupling to the computer system after all of the solution cycles have been computed.

10. A system for simulating an impact event in a finite element analysis used for assisting users to design or improve one or more structures, the system comprising:

a main memory for storing computer readable code for a finite element analysis application module;

a plurality of processing units coupled to the main memory, said processing units executing the computer readable code in the main memory to cause the finite element analysis application module to perform operations by a method of;

receiving a finite element analysis model of one or more structures subject to an impact event, wherein the finite element analysis model contains a plurality of contact interfaces;

dividing a finite element analysis model into a plurality of domains containing at least a first domain and a second domain, the first domain is associated with a first processing unit while the second domain is associated with a second processing unit of the processing units;

determining a group-able correlation between each of the contact interfaces, and one or more of the domains;

performing a time-marching simulation of the impact event having a plurality of solution cycles;

at each of the solution cycles,

creating and sending a first data communication message from the first processing unit to the second processing unit in accordance with the group-able correlation such that maximum possible data communications relevant to the contact interfaces are exchanged;

performing a plurality of computations for the respective contact interfaces of the finite element analysis at each of the first and second processing units with data received in the first data communication message;

creating and sending a second data communication message in a reverse direction of the first data communication message, the second message is configured to transmit data resulted from the computations; and

displaying results of the simulation in a monitor coupling to the system after all of the solution cycles have been computed.

11. The system of claim 10, wherein the computer system contains at least two processing units.

12. The system of claim 11, wherein the contact interfaces comprises at least two interfaces.

13. The system of claim 10, wherein the computer system comprises at least one computer, each having more than one central processor.

14. The system of claim 10, wherein the computer system contains more than one computer, each having at least one central processor.

15. The system of claim 14, wherein the central processor contains at least one processing core.